Isotope-based Fluvial Organic Carbon (ISOFLOC) Model: Model formulation, sensitivity, and evaluation
Abstract
Watershed-scale carbon budgets remain poorly understood, in part due to inadequate simulation tools to assess in-stream carbon fate and transport. A new numerical model termed ISOtope-based FLuvial Organic Carbon (ISOFLOC) is formulated to simulate the fluvial organic carbon budget in watersheds where hydrologic, sediment transport, and biogeochemical processes are coupled to control benthic and transported carbon composition and flux. One ISOFLOC innovation is the formulation of new stable carbon isotope model subroutines that include isotope fractionation processes in order to estimate carbon isotope source, fate, and transport. A second innovation is the coupling of transfers between carbon pools, including algal particulate organic carbon, fine particulate and dissolved organic carbon, and particulate and dissolved inorganic carbon, to simulate the carbon cycle in a comprehensive manner beyond that of existing watershed water quality models. ISOFLOC was tested and verified in a low-gradient, agriculturally impacted stream. Results of a global sensitivity analysis suggested the isotope response variable had unique sensitivity to the coupled interaction between fluvial shear resistance of algal biomass and the concentration of dissolved inorganic carbon. Model calibration and validation suggested good agreement at event, seasonal, and annual timescales. Multiobjective uncertainty analysis suggested inclusion of the carbon stable isotope routine reduced uncertainty by 80% for algal particulate organic carbon flux estimates.
Document Type
Article
Publication Date
6-2015
Digital Object Identifier (DOI)
http://dx.doi.org/10.1002/2015WR016999
Funding Information
We thank the University of Kentucky Department of Civil Engineering for partial funding of the graduate student. We gratefully acknowledge financial support of this research under National Science Foundation Award 0918856 and Kentucky Science & Engineering Foundation Award 2687-RDE-015.
Repository Citation
Ford, William I. and Fox, James F., "Isotope-based Fluvial Organic Carbon (ISOFLOC) Model: Model formulation, sensitivity, and evaluation" (2015). Civil Engineering Faculty Publications. 7.
https://uknowledge.uky.edu/ce_facpub/7
Figure 1 Image: Reach-scale conceptual model of the fluvial carbon cycle in low-order streams including carbon pools and the physical, chemical, and biological processes impacting their composition.
wrcr21495-fig-0001_10.1002%2F2015WR016999.pptx (195 kB)
Figure 1 Powerpoint: Reach-scale conceptual model of the fluvial carbon cycle in low-order streams including carbon pools and the physical, chemical, and biological processes impacting their composition.
image_n_wrcr21495-fig-0002.png (58 kB)
Figure 2 Image: Model flowchart for ISOFLOC which details inputs, outputs, and calibration procedures. The flowchart provides the framework for coupling hydrologic, hydraulic, sediment transport, organic carbon, and carbon isotope subroutines.
wrcr21495-fig-0002_10.1002%2F2015WR016999.pptx (210 kB)
Figure 2 Powerpoint: Model flowchart for ISOFLOC which details inputs, outputs, and calibration procedures. The flowchart provides the framework for coupling hydrologic, hydraulic, sediment transport, organic carbon, and carbon isotope subroutines.
image_n_wrcr21495-fig-0003.png (58 kB)
Figure 3 Image: Model evaluation procedure in ISOFLOC for (1) exploratory sensitivity analysis, (2) model calibration/validation, (3) uncertainty analysis, and (4) validation of output through comparisons with similar watershed systems.
wrcr21495-fig-0003_10.1002%2F2015WR016999.pptx (210 kB)
Figure 3 Powerpoint: Model evaluation procedure in ISOFLOC for (1) exploratory sensitivity analysis, (2) model calibration/validation, (3) uncertainty analysis, and (4) validation of output through comparisons with similar watershed systems.
image_n_wrcr21495-fig-0004.png (50 kB)
FIgure 4 Image: Modeling domain for the main stem of the South Elkhorn watershed (box) and its associated location within the HUC 6 Kentucky River Basin (bottom right). The polygons within the Kentucky River Basin represent HUC 12 watersheds.
wrcr21495-fig-0004_10.1002%2F2015WR016999.pptx (201 kB)
Figure 4 Powerpoint: Modeling domain for the main stem of the South Elkhorn watershed (box) and its associated location within the HUC 6 Kentucky River Basin (bottom right). The polygons within the Kentucky River Basin represent HUC 12 watersheds.
image_n_wrcr21495-fig-0005.png (150 kB)
Figure 5 Image: Sensitivity analysis of the ISOFLOC model displays the response of (left) δ13 CFPOC-T(av) and (right) CFPOC-T(av) to the eight most sensitive model parameters. The values plotted the average specified parameter value for that run.
wrcr21495-fig-0005_10.1002%2F2015WR016999.pptx (302 kB)
Figure 5 Powerpoint: Sensitivity analysis of the ISOFLOC model displays the response of (left) δ13 CFPOC-T(av) and (right) CFPOC-T(av) to the eight most sensitive model parameters. The values plotted the average specified parameter value for that run.
image_n_wrcr21495-fig-0006.png (37 kB)
Figure 6 Image: Seasonally averaged uncertainty analysis using the model evaluation framework with (a, b) carbon content of transported fine sediments and (c, d) both carbon content of transported fine sediment and stable carbon isotopic signature of transported fine sediment.
wrcr21495-fig-0006_10.1002%2F2015WR016999.pptx (189 kB)
Figure 6 Powerpoint: Seasonally averaged uncertainty analysis using the model evaluation framework with (a, b) carbon content of transported fine sediments and (c, d) both carbon content of transported fine sediment and stable carbon isotopic signature of transported fine sediment.
image_n_wrcr21495-fig-0007.png (31 kB)
Figure 7 Image: Event variability for the (a, b) elemental and (c, d) isotope models. Plotted values are deviations between calibration points for measured (y axis) versus modeled (x axis). Points that plot in first and third quadrants indicate the model adequately captures between-event variability. Points that plot in the second and fourth quadrants indicate that the model does not capture between event variability.
wrcr21495-fig-0007_10.1002%2F2015WR016999.pptx (183 kB)
Figure 7 Powerpoint: Event variability for the (a, b) elemental and (c, d) isotope models. Plotted values are deviations between calibration points for measured (y axis) versus modeled (x axis). Points that plot in first and third quadrants indicate the model adequately captures between-event variability. Points that plot in the second and fourth quadrants indicate that the model does not capture between event variability.
Notes/Citation Information
Published in Water Resources Research, v. 51, no. 6, p. 4046-4064.
© 2015. American Geophysical Union. All Rights Reserved.
The copyright holders have granted the permission for posting the article here.